In this Comment it is argued that Stishov et al. [Phys. Rev. B 86, 064433 (2012)] incorrectly estimated concentrations of (supposed) paramagnetic centers with \mu_eff = 4.8\mu_B in the investigated CoSi crystals. Correct estimation gives concentrations of such centers from 25 to 50 times smaller than reported (appr. 0.04 to 0.16% instead of appr. 2 to 4%). Also the reported data on temperature dependence of resistivity \rho(T) of four CoSi crystals prepared in different Labs are so close to each other at T = 250 to 300K that it is extremely unlikely to be reproducible for any reasonable accuracy of resistivity measurements. These and some other problems of the paper are related to the key points of the authors argumentation. As a result their main conclusions become unjustified.